Novel bHLH and WD40 transcription factors from turmeric (Curcuma longa L.) as putative regulators of curcumin biosynthesis

Turmeric, the golden spice belonging to the family Zingiberaceae, is enriched with biologically active curcuminoids composed of curcumin, demethoxycurcumin and bisdemethoxycurcumin. Curcuminoids are phenylpropanoid derivatives, and the biosynthetic pathway is controlled by several transcription factors (TFs). bHLH, WD40 and MYB TFs are the most important TFs regulating phenylpropanoid biosynthesis in plants. Through comparative transcriptome analysis of high and low curcumin germplasm accessions, 20 TFs belonging to the classes bHLH, WD 40, NAC, WRKY and bZIP, which showed differential expression with respect to curcumin, were identified. Among these, two bHLH and one WD40 TFs showed maximum comparative fold change and negative correlation vis-a-vis curcumin content in quantitative real-time polymerase chain reaction (qRTPCR) analysis. The results of comparative transcriptome and qRT-PCR analyses were in congruence, indicating their putative role as negative regulators.


Introduction
Curcuminoids are biologically active compounds present in rhizomes of turmeric (Curcuma longa L.). Curcuminoids are a mixture of curcumin, demethoxycurcumin and bisdemethoxycurcumin. In recent years curcumin has attracted significant attention due to its health-promoting effects, such as protection against cancer, heart diseases (Wongcharoen and Phrommintikul, 2009) and other chronic human disorders (Mantzorou et al., 2018). Moreover, curcumin has also been used as a natural replacement for synthetic food colourants (Ismail and Sakr, 2016) and is an integral part of religious ceremonies (Chattopadhyay et al., 2004).
Curcuminoids are a class of secondary metabolites synthesised through complex biosynthetic pathways that involve structural genes responsible for the formation of final metabolite and regulatory genes that regulate their expression.
Curcuminoid biosynthesis involves several enzymecoding structural genes such as phenylalanine ammonia lyase (PAL), cinnamate 4 hydroxylase (C4H), 4-coumaryl CoA ligase (4CL), diketide CoA synthase (DCS) and curcumin synthases 1, 2, and 3 (CURS1, CURS2, CURS3) (Katsuyama et al., 2009a;2009b). Earlier studies from our lab have reported certain novel structural genes like CLPKS11 with a putative role in biosynthesis (Deepa et al., 2017). The expression of these structural genes is regulated at the transcriptional level by TFs. TFs are sequencespecific DNA-binding proteins that interact with the regulatory regions of the target genes and control transcriptional initiation rate (Vom Endt et al., 2002). These regulators bind to homologous cis-elements situated in promoters and regulate the expression of biosynthetic pathway genes or interact with other TFs to form a complex , Zhao et al., 2019. Several families of TFs have been described as regulators of plant secondary metabolism that include bHLH, WD40, MYB, WRKY, bZIP, NAC etc. (Zhao et al., 2013;Saga et al., 2012;Xu et al., 2004;Wang et al., 2015). Biosynthesis of phenylpropanoids is generally controlled by a gene regulatory network involving MYB transcription factors regulating early biosynthetic steps and a complex of MYB-bHLH-WD40 TFs regulating the late biosynthetic steps (Li et al., 2014). Late biosynthetic steps are more specific to the final metabolite and are more important for pathway engineering. We have already reported MYB TFs regulating curcumin biosynthesis in an earlier study (Sheeja et al., 2015); however, no information on key TFs such as bHLH and WD40 are available. Hence, in the present study, we evaluated bHLH and WD40 TFs and identified three novel TFs that negatively regulate curcumin biosynthesis, which may lead to unravelling the regulatory mechanism.

Plant material
Rhizomes were collected after 120 days of planting from high-curcumin turmeric-IISR Prathibha, low curcumin turmeric-Accession No. 200 and a closely related species that is practically devoid of curcumin-C. aromatica from the Experimental Farm, Chelavoor of ICAR-Indian Institute of Spices Research (ICAR-IISR), Kozhikode, Kerala State, India (Table 1).

Screening of Curcuma transcriptome for different transcription factor unigenes
Homologous transcripts of different types of TFs like bHLH, WD40, NAC, WRKY, bZIP with fold change >3 was retrieved from the text of annotated genes within transcriptome data of C. longa (high curcumin) and C. aromatica (practically devoid of curcumin) (Sheeja et al., 2015) (Table 2). Gene ontology (GO) terms were retrieved for each unigene using the default settings of Blast 2GO software (Conesa and Gotz, 2008).

Quantitative real-time PCR
Total RNA was isolated as per Deepa et al. (2014), and three biological replicates were pooled for each sample and were digested using an oncolumn DNase I digestion set. The quality and quantity of RNA were checked by agarose gel electrophoresis (1.2%) and Biophotometer plus (Eppendorf, Germany). The RNA samples showed two discrete bands of 28S and 18S rRNA on agarose gel and A260/280 ratio between 1.8 and 2.0. About 250 ng of total RNA was used for first-strand cDNA synthesis using Superscript III reverse transcriptase (Invitrogen) and Oligo-(dT)18 primer in a total volume of 20 μL. Gene expression profiling was performed on Rotor-Gene Q (Qiagen) using the QuantiFast SYBR Green PCR kit (Qiagen). The reaction mixture comprised of 10 μL of 2X SYBR Green, 10 pmol each of gene-specific primers and 100 ng of cDNA in a final volume of 20 μL. qRT-PCR amplification was performed under the following conditions: 95 o C for 5 min, followed by 40 cycles of 95 o C for 10s and 60 o C for 30 sec. A melt curve program of 65-99 o C was included to check the specificity of qRT-PCR products. Three technical replicates per reaction were performed, and the mean Ct value was used for the analysis. A reverse transcription negative control (without reverse transcriptase) and a non-template negative control were included to confirm the absence of genomic DNA and check non-specific amplification. The primers for 20 TFs were designed based on previous transcriptome data (Sheeja et al., 2015) using Primer Quest (www.idtdna.com/primer quest/ home/index); the details of the primers are listed in Table 2. Candidate reference genes viz., EF1α (FP-GCTGACTGTGCTGTTCTCATTAT and RP-CTCGTGTCTGTCCATCCTTTGAA) and Ubiquitin (FP-GCACTCTCGCTGACTACAAC and RP-GGCTTGGTGTAGGTCTTCTTC) were used as per Deepa et al. (2017). Standard curves were generated for each gene with serial dilutions of pooled cDNAs (10 -1 to 10 -6 ) for calculating qRT-PCR efficiency. The relative gene expression was calculated using the 2 −ΔΔCt method (Livak and Schmittgen 2001) and transformed to a log 2 scale.

Estimation of curcumin content
Curcumin content of rhizome samples was analysed spectrophotometrically at 430 nm following the American Spice Trade Association (1968) procedure (Table 1).

qRT-PCR expression patterns of TFs
qRT-PCR analysis is a common strategy to validate the RNA-Seq data. Twenty shortlisted TFs from transcriptome analysis were further evaluated by qRT-PCR with IISR Prathibha as control. Among these contig IDs 603014 and 668415 (bHLH TFs) and contig ID 677063 (WD40 TF) showed the highest relative fold change with respect to curcumin (low vs high) (Fig. 1). The expression of ClPKS11, a bait gene that is found to be upregulated under high curcumin conditions (Deepa et al., 2017), was high in IISR Prathibha and down-regulated in low curcumin accessions of C. aromatica and Accession No. 200. bHLH TFs (Contig IDs 603014 and 668415) and WD40 TF (contig ID 677063) showed a negative correlation with respect to the expression of ClPKS11 (Fig. 1). If one or more genes in a biosynthetic pathway for a particular metabolite is identified, these genes may be used as baits to identify other novel candidates involved in the pathway through a co-expression analysis (Medema and Osbourn, 2016). Several studies have reported that bHLH members play a central role in transcriptional regulation of biosynthetic pathways by forming a complex with WD40 and MYB TF proteins (Zhao et al., 2019). In flavonoid biosynthesis, additional regulators which interact with bHLHs or R2R3-MYBs to organize or disrupt the formation of the MBW complex, leading to enhanced or compromised flavonoid production, have been reported (Li et al., 2014). The qRT-PCR trends were in congruence with the fold change values in RNA-seq Data. Thus, we have identified three novel TFs belonging to the bHLH and WD40 classes to add up to the several MYB TFs identified in our earlier studies (Sheeja et al., 2015). Blastn search indicated the unique status of all three TFs based on the per cent identity (Table 2). These novel TFs, in conjunction with the already identified MYB TFs will be instrumental in future studies to confirm the mechanism of the molecular regulation of curcumin biosynthesis.  Table 2, the arrow represents the novel candidate TFs identified by qRT-PCR.

Conclusion
Curcumin is a phenylpropanoid derivative, and its biosynthesis is regulated by several classes of TFs, with bHLH, WD40 and MYB TFs being the most important. In the present study, we have identified 20 TFs through comparative transcriptome analysis. Among these, real-time expression patterns of two TFs belonging to bHLH and one belonging to WD40 categories indicated the highest differential fold change between high and low curcumin accessions and a negative correlation with curcumin content. Their expression was also negatively correlated with a bait gene CLPKS 11 in both transcriptome analysis and qRT-PCR. These novel TFs may help understand the molecular regulation of biosynthesis of curcumin, which is essential for developing high curcumin lines in turmeric.